CIA-RDP96-00789R002900090001-1

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
SECRET
DT-S-1008-5
,
DIA
DEFENSE
INTELLIGENCE
AGENCY
ADP SYSTEM REQUIREMENTS (U)
Initial Report
20 December, 1990
NOFORN
SECRET
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
-
SECRET
ADP SYSTEM REQUIREMENTS (U)
INITIAL REPORT
:
..
Date of Publication
15 December, 1990
This is a Department of Defense Intelligence Document
prepared by the Technology Assessment and Support Office,
Directorate for Scientific and Technical Intelligence
Defense Intelligence Agency
SG1J
PREPARED BY
for Technology Assessment and Support Office (DT-S)
NOT RELEASABLE TO FOREIGN NATIONALS
CLASS BY: DIA/DT
DECLASS:
OADR
-
..
SECRET
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1

4
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
-
ADP SYSTEM REQUIREMENTS (U)
INITIAL REPORT
I
TABLE OF CONTENTS
EXECUTIVE SUMMARY
i
I.
PURPOSE
1
II.
SCOPE
III. THE SINGLE MOST IMPORTANT ACTION
.
1
IV.
EVENTS AND AN ESTIMATE OF THE TIME REQUIRED
2
V.
PERT CHART
5
VI
PRELIMINARY HARDWARE/SOFTWARE ESTIMATES
APPENDIX 1:
MAJOR CONSIDERATIONS FOR EACH STEP
7
22
APPENDIX 3:
EXAMPLES OF SURVEY RESULTS (U)
-
:
-
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
6
SG1A
1

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
-
-
EXECUTIVE SUMMARY
1.
(U) This report presents the results of a preliminary analysis of
the project's ADP needs, a recommended method of system acquisition,
preliminary schedule/cost estimate, and supporting analytical data.
2.
(C)
Based on the preliminary analysis, it is estimated that
approximately $80,000.00 to $116,000.00 would be needed to acquire the
hardware/software necessary to meet the program's ADP needs. Some of the
major items identified are:
a.
New computers (80386-based systems to meet increased needs).
The current system (80286-based) cannot efficiently support the
large databases anticipated, nor will they be compatible with
anticipated contractor's systems.
b.
Text and graphic digitizers for data input.
c.
SYBASE or comparable database software package.
d.
Standard software (word processing, spreadsheet, etc. ).
e.
Research input devices (biofeedback, EEG, etc. )
f.
Laser printer (s) for professional output.
g·
Projection monitor for presentation.
h.
Telecommunications hardware/software.,
3.
(U) Since the program facilities are physically separated from the
main DIA technical facilities, most of the acquisition effort will have
to be performed by program personnel. At present, there is only one
program member who has the requisite computer skills (a background in
systems analysis and design) to manage the acquisition process. The
estimates in this report are based on the Computer Operations Manager
dedicating an estimated 40% of his time to the effort.
4.
(U) Based on this limitation it will require approximately 12
months to achieve system Initial Operating Capability. This estimate
could be reduced by up to 6 months, if additional qualified personnel
were assigned to assist in the development activity.
-
5.
(U) Due to this long lead time, it is requested that management
review/approval of the proposed approach be granted as soon as possible.
i
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
-
-
*
-
ADP SYSTEM REQUIREMENTS (U)
INITIAL REPORT
I.
(U)
PURPOSE:
(U) The purpose of this initial report is to delineate the steps
necessary for the development/upgrading of the unit's ADP system to
adequately meet present and future needs. More detailed aspects of each
portion of development will be issued in follow-on reports.
II.
SCOPE:
(U) This report presents an analysis and PERT chart of the work
that must be performed in order to acquire an ADP system to meet this
project's administrative, training, R&D, and database management needs,
with special attention to the management of the Foreign Intelligence
database and other topic-specific databases.
III. (U) THE SINGLE MOST IMPORTANT ACTION:
(U) The most important aspect of procuring/upgrading an ADP system
is planning. Millions of government dollars have been wasted on new
equipment which, after payment and delivery, is not compatible with the
old system, does not meet the users' real needs, or is so "user
unfriendly" as to actually hinder efficient office operation. The only
preventive measure for this is proper planning.
(U) For example, many, if not most, computer systems are bought
because a computer vendor demonstrates impressive capabilities, or
because the office wants to buy the "most powerful" system for the amount
of money allocated in the budget. The most common result is that the
system users have to change their work habits to meet the new system's
shortcomings and idiosyncracies. They often learn too late that although
the system may be good for engineering applications, it is terrible at
word processing or database searches, which may constitute the bulk of
the computer's real work load.
(U) The proper way to plan a system is to learn first what you
need, not what a vendor has to sell. Once you know what you need, then
find a vendor whose products can meet those needs. The user should not
have to meet the system's needs; the system should meet the user's needs.
program.
(U) These considerations take on even greater significance for this
Since it is physically separated from the main DIA technical
1
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
-

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
-
-
facilities, we do not have easy access to maintenance or technical
support. Such factors as maintainability, reliability, compatibility and
supportability, as well as the ability to function in a "stand-alone"
environment, must be given adequate attention up front in the planning
process, not only in terms of hardware/software, but also with respect to
program personnel qualified to operate/maintain/modify the system to meet
changing program requirements. At least one member of this project needs
to be a qualified Computer Operations Manager.
IV. (U) EVENTS AND AN ESTIMATE OF THE TIME REQUIRED:
(U)
Proper planning proceeds in a set pattern of specific steps.
The elimination of any of these steps, taking steps in the wrong order,
or taking shortcuts could spell disaster for the new system. The time
estimates shown reflect one person working part-time on the task.
A.
(U) TASK A: Establish the system's OUTPUT requirements:
(Survey of the unit's needs).
Figure 1 shows a list of the output
requirements at the time of the last system analysis for this office
(1985). An update of these requirements should take approximately 4-6
weeks.
B.
(U) TASK B: Establish the system's INPUT requirements (Survey
what is necessary to meet the unit's needs). This must include all
equipment, training, software, new work routines, etc. necessary for
allowing the data to get into the computer. Figure 2 lists just the
manhours necessary for keyboard entry, as of the last system analysis
(1985) This survey should not be started until the output requirements
are completely understood. It should take approximately 3-4 weeks after
the completion and approval of Task A.
c.
(U) TASK C: Determine operating and security requirements:
Such factors as operating environment, security regulations, etc. impact
directly on purchase considerations. If other personnel are available,
this step can be accomplished in the same timeframe as Task B. If not,
it will require an additional 3-4 weeks after the completion of Task B.
D.
(U) TASK D: Determine the system's THROUGHPUT requirements:
In other words, determine how much the system will have to handle at any
one time, as well as overall. This constitutes the system's working
parameters, and is "must know" information at purchase time. Figure 3
shows this information as of the last system analysis (1985). Estimated
time for determination and report: 2 weeks after the approval of Tasks B
and C.
2
-
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1

-
--
-
-
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
E.
(U) TASK E.
Determine other office impacts:
How will meeting
the present needs affect future needs? Only by determining what the
probable future needs will be can you make certain that the system is
capable of growing along with the unit, and will not become insufficient
to the unit's needs within the foreseeable future. Figure 4 is a list of
only the filing requirements necessary for the system as of the last
analysis (1985). This step should be a constant consideration during all
of the above surveys, but should be addressed separately after completion
of Task D, and should generate a report of its own. Estimated time for
completion: 2 weeks after the completion of Tasks A-D.
F.
(U)
TASK F. Conduct a market survey the possible
hardware/software available:
In other words, "shop around". This step
is most important, and shortcuts should not be taken. It is time-
consuming to talk to vendors, wait for information and pricing, etc. For
this reason, an estimated 3-4 weeks should be allotted for this step.
NOTE:
Maintenance requirements and repair costs must be considered in
this step as well as initial costs.
G.
(U) TASK G. Establish and document the final budget aspects.
This step can be performed at any time, but the above findings may call
for a final reconsideration and adjustment of the allotted amounts. If
amounts need to be renegotiated, this step can take 1-8 weeks.
H.
(U)
TASK H. Delineate contractual requirements and demands:
This step should be performed early in the process, as it may impact
drastically on all future decisions. Estimated time: 2 weeks, performed
at the same time as other steps.
I.
(U) TASK I. Order new hard- and software:
Time estimated for
writing purchase requests, purchase orders, etc. is 2-3 weeks.
J.
(U) TASK J. Decide which data and documents are to be
transferred to the new system: This task is performed by management, and
can be performed at any time.
K.
(U) TASK K. Delivery and installation:
Time requirements for
this step depend on the vendor, the paper-trail system, and a multitude
of unforeseeable factors. An estimated time frame is 4-8 weeks.
L.
(U)
TASK L. Installation and checkout of the system:
This
includes thorough testing of the system, using office data and documents.
During this time the Computer Operations Manager gains a thorough
knowledge of the system and the software, in order to train/help others.
3
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
-
.
1
-

4
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
+
-
:
1
-
Initial testing should take approximately 1 week.
M.
(U)
TASK M. Transfer and conversion of data and documents
from the old system to the new/ Acquisition of data and loading of
databases to provide an initial operating capability: Estimated time of
transfer for our office: 0 - 3 weeks. NOTE: This time estimate can vary
greatly according to the compatibility of the new software with the old.
N.
(U)
TASK N. Personnel training:
Initial training of
personnel prior to first usage of the system should require 1-2 weeks,
according to the software purchased. However, there is a period of up to
3 months afterwards while the users are still below maximum proficiency.
During this time, the System Administrator is called on quite frequently
to spend time working out problems with individual users.
4
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
V. (U) PERT CHART
(U) The following PERT chart shows the necessary scheduling for
tasks A-N.
-
Ł
PROJECT: ADP DEVELOPEMENT PLAN
ADP DEVELOPEMENT PLAN
11991
(Dal Who | Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Decl
Establish Sys Output Req
: OIX:
Establish Sys Input Req
1 01X
Determine Throughput Req
: 01X
Determine Ops/Security Req
: 01X
Determine Office Impacts
Establish Budget
: 01X
Market Survey (H/W-S/W)
: 01X
Determine Contract Req
: 01X
Order H/W-S/W
: 01X
Deliver/Install/Training
: 01X
Data Acq/DB Construction
: 01X
Initial Operating Capability: 01X
:
-
5
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
-
.
.
.

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
+
.
-
VI. (U)
PRELIMINARY HARDWARE/SOFTWARE ESTIMATES:
Although no exact costs can be given until the completion of Task F,
a rough estimate of system costs can be projected at this time. Wherever
applicable, 2 years of maintenance costs have been figured into the
estimated cost.
80386-based system (10 terminals)
$ 71,900.00
Laser printers (2)
16,000.00
Graphics digitizer
3,000.00
Optical character reader (OCR)
1,000.00
Software
7,000.00
Local Area Network (may not be needed)
25,000.00
Modem
(may not be needed)
200.00
Specialized research equipment
Biofeedback apparatus
500.00
Synergizer
500.00
EEG
800.00
Random Number Generator
100.00
TOTAL ESTIMATED COSTS
$116,000.00
-
6
C
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1

.
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
+
-
-
-
-
APPENDIX 1:
MAJOR CONSIDERATIONS FOR
EACH STEP
7
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
CRANE
(U)
Each step requires specific considerations.
Some considerations
already delineated by the date of this publication are:
A.
(U)
First, establish the system's OUTPUT requirements: Some
important considerations are:
1.
(U)
Printed material requirements: Printed matter in the
form of reports, database summaries, etc. makes up the bulk of an
office's computer usage. It is the most visible and long-lasting output
the system will make. It is therefore the most important Output
consideration. Report types, quantity and quality of print, single vs.
multiple sheet printing, etc. must be considered in order to build a
system which will meet the unit's needs. Figure 1 contains a listing of
the printed output requirements this office had when the last systems
analysis was performed. A more up-to-date listing will accompany a more
detailed report. There are three basic forms of printers:
a.
(U) "Daisy wheel" (also called "letter quality, even
though all three types can produce letter-quality print"). This printer
has a rotating wheel which strikes the paper, leaving an impression of
the letters. It is slow, loud, has very little flexibility for
specialized purposes. It does not have any graphics capability.
b.
(U) "Dot Matrix" (including "ink jet"). This
printer has a "print head", containing small pins which strike out in the
required patterns to print text or graphics. It is more compact, less
noisy, and very flexible for specialized purposes. In fast mode, the
letters and graphics are made of coarse dots. In slow mode, the dots can
barely be seen, but the slow mode is very time-consuming, and usually
keeps the user from any other activity on the computer while printing
takes place. Printed output, especially graphics, is usually of less
quality than desired.
c.
(U)
"Laser". This printer uses a laser to form an
image on a xerox drum which is then transferred to the paper. It is
fast, quiet, does text and graphics with extremely professional looking
results. However, it will not do carbon copies,
2.
(U)
Graphics output requirements:
a.
(U) With the modern trend toward presentation in
graphic format rather than as printed text, graphics requirements are a
major concern. Figure 1 includes the graphics requirements, since almost
all graphics are printed before being changed to other presentation
formats.
8
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
-
.
-
-
I
-

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
b.
(U) Another form of graphics output is that of the
projection-type computer monitor. This adds both color and motion impact
to presentations and to computerized training.
There are two basic forms
of projection equipment for this purpose:
1)
The Overhead Projector Slide Screen:
This is a
monochrome monitor which is the same shape and size of an overhead
projector slide. It hooks to a nearby computer, and forms an image
exactly the same as that on the computer screen. When this is laid on a
standard overhead projector, the screen image is cast onto the wall or
projection screen, just like any standard overhead projector slide.
This
allows the presentation to be controlled by the computer operator. The
cost of this type of monitor is approximately $400, and is limited to
monochrome displays.
2)
A dedicated projection monitor. This type
monitor has a standard color monitor with a lens in front of it which
projects the computer image onto the wall or a projection screen.
The
cost is in the $800 range, more or less, depending on quality, projection
distance, color quality, etc. Like the other type, this does not require
a dedicated computer, but simply hooks to a standard machine for
projected output.
3.
(U)
Database presentation requirements
a.
Will output be merged into other output or stand
alone?
b.
Will data output require accompanying graphics?
c.
Will database software be compatible with other
software being used on the system?
d.
What other equipment is necessary for output?
1)
Printer or plotter
2)
Modem (for tele-output) - does this need to be a
secure line with crypto?
3)
Projection screen terminal (for briefings)
4)
Color vs. monochrome
-
9
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1

-
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
.
.
B.
(U) Second, establish the system's INPUT requirements: Having
established the OUTPUT requirements shows the planner what INPUT hard-
and software is needed. For example, if maps are to be included in
presentation graphics as part of the output, maps will need to be INPUT.
There is only certain equipment which can digitize maps and put them into
the computer, and the planner is automatically pointed to the need for
such equipment.
Beyond the obvious input requirements, however, there are other
considerations:
1.
(U)
How labor intensive is the input? Is there equipment
on the market which will be cost effective, due simply to a savings in
man-hours saved?
2.
(U) How user friendly is system? No matter what the
system costs, it costs too much if the user will not use it.
3.
(U) What other equipment is necessary for input?
The
basic system rarely ever has everything you need for any work which is
more advanced that just text or data input. For the program's purposes,
the following input devices need to be evaluated:
a.
(U) Graphics digitizer: This input device is much
like a xerox machine, but instead of a copy of the original, will
digitize pictures, sketches, artwork, etc. into a format which can be
incorporated into word processing documents, briefing slides, etc. This
is a graphics only device.
b. (U) Optical Character Reader: This input device is
also like a xerox machine, but will read a document and input the text
into word processing documents. This is a text only device.
c.
(U) Modem: This input/output device connects the
system, through phone lines, to other computer systems, databases, etc.
It allows the system to draw off data and information from other sources,
send documents to other sources, etc. Secure modems may or may not also
require:
1)
A Crypto device for encipherment of the data,
documents, etc.
2)
A secure telephone line for transmission.
d.
(U)
Special input devices specific to the program's
10
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
..
-
mission:
Such input devices are bought separately from the normal
equipment we would buy for its regular mission. However, this
requirement on the program's computer system means that, when those other
pieces of equipment are bought, they will have digital outputs which can
be read by the computer (they usually do not have). Such specialized
devices include, but are not limited to:
1)
Galvanic Skin Response (GSR) device:
This
bottom-of-the-line device measures the minor variations in the skin's
resistance to an electrical microcurrent. Such variations indicate the
worker's mental and emotional state at any given time.
The machine,
therefore, works as a monitoring device. If such a machine were used, it
should have a port for output of the results in digital form.
2)
Biofeedback: Much like the GSR, a biofeedback
machine allows the monitoring of the worker's mental and emotional
states, through the monitoring of brainwave activity.
This device has
the added ability to give the worker audible feedback. Therefore, when
the worker's mental state is less than desirable, he/she will know, and
can immediately work to bring themselves back into optimum working order.
Again, such a machine can only provide an input to the statistical
databases if there is a port for the information in digital form.
3)
EEG: An electroencephalograph (EEG) machine has
never been used for this office, but the need for one has long been
recognized. It is a more sensitive machine than the biofeedback machine,
and can keep track of much more data. It does not provide feedback,
unless used with the hemisync device, mentioned in the next item, below.
For such use, a digital port is required.
4)
Hemisync: This device uses tones to help attain
working states in a much shorter time, as well as to help train new
personnel in attaining these ideal working states. The standard machine
of this type merely produces tones. However, with digital input/output,
the hemisync machine can work in conjunction with the EEG and the
computer. The computer, while recording the statistics involved in the
work period, can use the information coming from the EEG to control the
tonal output of the hemisync machine, to produce the optimum session
working-state parameters. This EEG-hemisync combination is the optimum
setup presently available for the program's location and exclusive use.
5)
Magnetoencephalograph:
This device is
exponentially more accurate than the EEG and can be used in much the same
way as the EEG-hemisync combination. The cost, however, is totally
prohibitive, requiring special cryogenic devices, a special facility,
11
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
etc. There is no justification for having such a machine for the
program's own private use. There is a proposal that workers might go to
a place where one of these machines is located, strictly for research
purposes. If so, plans should be made to secure special tapes or disks
which contain the digital results, for use in the program's statistical
studies.
C.
(U)
Determine special operating and security requirements:
One set of special requirements is addressed in Appendix 2 (Foreign
Intelligence Database). Other major considerations are:
1.
(U)
Personnel required for:
a.
System management
b.
Software maintenance
c.
Database administration
d.
Software installation/checkout
e.
Identification of new software/hardware needs as
office growth increases.
f.
Recovery of lost/damaged files/data.
2.
(U) Programming languages for special operating needs:
Can off-the-shelf software suffice for all the office's needs? If not,
there are three options for specialized software:
a.
(U) Contract for the special software needed. This
option is usually a non-option, due to its expense (up to
$50,000/program) in both time and money. It also impacts on security,
since an outside contractor has to be read onto any secure information
which will be manipulated by his program.
b.
(U) Request special programming from DIA. This
option is costly in respect to time. If the program is very complex, it
may still be contracted out.
c.
(U) Include a programming language in with the
software ($40 - $100), and have or train already cleared in-house
personnel to use it. This is the preferred and most time and money
effective method. At present, the program has only one person who is
fully qualified to program specific needs. Good planning, however, might
12
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
.

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
.
.
include a requirement in the Personnel Selection criteria to insure that
at least one new incoming personnel has programming ability.
3.
(U) Should workstations be interconnected? (Stand-alone
vs. LAN or mini-computer).
There are too many pros and cons for each
decision to be addressed in a document of this size. The determination
makes minor differences in system capabilities, but makes huge
differences in terms of system costs and flexibility. A major
consideration, however, would be whether or not everyone on the program
needs on-line access to the databases. This, of course, is a management
issue.
SG1A
4.
(U)
Does the program need a Tempest system?
In light of
the program's current facility and anticipated database classification
level, security requirements must be studied in detail. Tempest-secure
systems have no more capabilities than regular systems, and normally have
little impact on the user.
The costs of such a system, however, are
sometimes doubled or tripled.
When judging this aspect of system needs,
one must take into account the facility in which the system is located,
and all other security aspects. One obvious method of cutting the cost
is to determine whether or not Tempest approval is needed on all
terminals, or whether one or two dedicated Tempest terminals will
suffice. Also, some means of controlling the perimeter around the
program's facilities might be feasible to meet this requirement.
6.
(U) Will the system need data transmission capabilities?
This consideration impacts directly on all decisions about input, output,
and security.
7.
(U)
Will additional training be required
a.
Because the system imposes new security risks?
(Specifically, in this case, how will the handling of
SI/TK Codeword material impact operations?)
b.
For the System Administrator?
c.
For the users?
8.
(U)
How will in-house research impact on the needs of the
5
13
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
-

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
L
-
system? Several pieces of equipment will be bought in other purchases
for in-house research, such as EEG machines, biofeedback, Random Number
Generators, etc. In designing the system, it must be remembered that
these machines will probably have digital ports to feed data directly
into the computer, and/or for direct computer control. This stage of
planning must take into account that INPUT and OUTPUT ports will be
needed to access these machines.
D.
(U)
Determine maintenance requirements and repair costs: It
is often true that maintenance winds up costing more than the equipment,
itself. Major considerations in this area are:
1.
(U) Will DIA's present maintenance contract cover new
equipment?
2.
(U)
Will some of the equipment come with its own
warranties, alleviating the need for immediate service contracts and/or
saving on contractual costs for maintenance?
E. (U) Determine other office impact: Several of the specific
projects which are to be performed will impact the office environment in
the form of work schedules, deadlines, etc.
F.
(U) Determine future growth needs and uses: A system which is
designed to meet present needs only will tend to be obsolete the day it
is installed. If the machine actually does do its job of allowing the
office to become more productive and meet present needs in a shorter
time, its very presence in the office will cause growth. If it cannot
meet this growth, the fault is that of the planner. This stage of the
planning is perhaps the most nebulous. No one can accurately predict
what future needs will arise. However, the planner must try to logically
look at the office's goals, personnel, and must consider ALL future plans
the office is making (not just those related to the system), to be
certain the system will meet future needs.
G.
(U) Once a complete understanding of the needed system is
achieved, conduct a market survey of hardware/software meeting the
requirements. The major considerations which must be addressed include,
but are not limited to:
1.
(U) Software/hardware compatibility.
2.
(U) Software/user compatibility.
3.
(U) CAN THE FILES FROM THE OLD SYSTEM BE CONVERTED TO THE
14
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
.
..
-
--
-
NEW SYSTEM?
If not, old documents and databases will have to be either
retyped, or will be lost to computer usage. This is unacceptable. This
aspect is usually ignored when lining up purchase plans, and almost
always results in a disastrous situation for the office.
H.
(U)
Make a survey of vendors, purchase resources, prices, etc.
1.
Is it necessary to go through DIA-established channels?
If so, what are they, and what information/actions are
required on the program's part?
2.
Can the program go directly to a contractor?
3.
How much off-the-shelf hard/software can be used? Will it
save money to do so?
I.
(U)
Delineate contractual requirements
1.
(U)
Is DIA approval needed for hard/software?
2.
(U)
Who handles the contracts?
3.
Matching funding to costs. How is overall money
:
distribution affected if one thing cannot be purchased or
if maintenance costs for one thing mean that something
else has to be given up? Is the overall picture
considered for EVERY change made?
J.
(U)
The actual mechanics of replacing the present system
1.
Are the users briefed and prepared for the inconveniences
facing them?
2.
Which is the best installation method: all at once or in
"builds" (gradual integration)? Each has its own time
demands on the users.
3.
Software installation on new system requires a period of
"check-out" before the users are proficient. How long
will this take?
4.
Transfer and/or conversion of data and document files,
loading of databases, rewriting of program macros, etc.
5.
(U)
Training/Re-training of users
15
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1

SG1A
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
Next 5 Page(s) In Document Exempt
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
.
-
...
.
APPENDIX 3:
EXAMPLES OF SURVEY RESULTS (U)
The examples in this appendix are taken from the last system analysis,
performed in 1985 on the Wang computer system which the program used at
that time. Some of the items will not apply to the present situation,
and some items which apply to the present situation will not be found in
these examples. The purpose of this section is simply to allow the
reader to better understand the level of detail which must be considered
(but is often ignored) in ADP system acquisition and to indicate the
minimum requirements that the proposed system will have to meet.
22
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
-
-

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
EXAMPLE 1:
(U)
MONTHLY DOCUMENT OUTPUT REQUIREMENTS
(Example report from 1985 System analysis - not current information)
TYPE OF DOCUMENT
#/mo. pgs/doc
pages/mo
Access Roster database
01
X
10
=
010
pages/mo
Briefing texts
02
X
20
040 pages/mo
Charts/diagrams
40
X
01
=
040 pages/mo
Cmd Operating Budget
01
X
02
=
002 pages/mo
Concept papers
01
X
20
=
020 pages/mo
Contract negotiations
01
X
50
050
pages/mo
Contract evaluations
01
X
15
=
=
020
pages/mo
Daily Training Summ.
20
X
04
080
pages/mo
Data manipulation*
02
X
03
006
pages/mo
Decision papers
14
X
07
098
pages/mo
Disposition forms
12
X
02
=
024 pages/mo
Fact sheets
03
X
08
024 pages/mo
File backup record
04
X
02
008 pages/mo
Files deleted report
01
02
=
=
=
=
=
=
008 pages/mo
Memos for record
18
X
02
=
036 pages/mo
Mileage report
01
X
01
001 page /mo
Monthly Training Rpt.
1
X
20
=
=
010 pages/mo
Outside correspondence
15
X
02
=
030 pages/mo
Personnel evaluations
03
X
02
=
012 pages/mo
Proj. Ofc. hist. rpt.
06
X
02
=
012 pages/mo
Project Officer Reports
05
X
10
=
050 pages/mo
Project summaries
05
X
05
=
=
=
=
040 pages/mo
Stage essays
01
X
05
=
005 pages/mo
Talking papers
02
X
12
=
024 pages/mo
Technical summaries
10
X
25
=
250 pages/mo
Training database
20
X
01
=
=
002 pages/mo
Travel reports
02
X
05
010 pages/mo
Utility assessments
01
X
09
=
009 pages/mo
VuGraph slides
05
X
01
=
005 pages/mo
Misc. other
25
X
01
=
025 pages/mo
TOTAL PRINTED OUTPUT/MONTH
= 1387 pages/mo
=
=
=
002 pages/mo
Historical report
01
X
10
010 pages/mo
Impact statements
01
X
08
008 pages/mo
Information papers
03
X
12
036 pages/mo
Inter-office corr.
80
X
03
240 pages/mo
Justification papers
01
X
08
020 pages/mo
Operations rqmnts. pkg.
05
X
02
025 pages/mo
Session results package
05
X
04
020 pages/mo
Session reports
60
X
04
240 pages/mo
Staff studies
02
X
20
020 pages/mo
Travel reimbursement
02
X
01
=
015 pages/mo
Cover letters
20
X
01
=
=
23
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
-
=
X

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
EXAMPLE 2: MANHOURS REQUIRED FOR KEYBOARD INPUT
(Example report from 1985 System analysis - not current information)
ITEM
#/mo. Hrs/doc
Hrs/mo
Monthly Training Rpt.
01
X
14
=
014 hrs/mo
Daily Training Summ.
20
X
03
=
060 hrs/mo
File backup record
04
X
01
=
004 hrs/mo
Files deleted report
01
X
02
=
002 hrs/mo
Operations rqmnts. pkg.
05
X
01
=
005 hrs/mo
Session transcripts
60
X
05
300 hrs/mo
Session reports
60
X
02
120 hrs/mo
Session results package
05
X
05
025 hrs/mo
Training database
20
X
01
=
020 hrs/mo
Technical summaries
01
X
05
005 hrs/mo
Inter-office corr.
80
X
02
=
=
=
004 hrs/mo
Travel reports
02
X
02
=
004 hrs/mo
Briefing texts
02
X
08
016 hrs/mo
Charts/diagrams
20
X
01
020 hrs/mo
Mileage report
01
X
01
=
001 hr /mo
Physical training rpt
01
X
01
=
001 hr /mo
Contract negotiations
01
X
50
=
050 hrs/mo
Contract evaluations
01
X
20
=
020 hrs/mo
Decision papers
14
X
12
=
168 hrs/mo
Fact sheets
03
X
03
=
009 hrs/mo
Information papers
02
X
12
=
=
005 hrs/mo
Proj. Ofc. hist. rpt.
06
X
02
=
012 hrs/mo
Personnel evaluations
03
X
03
=
009 hrs/mo
Incls/indorsmts/etc
20
X
01
=
=
003 hrs/mo
Staff studies
02
X
20
=
040 hrs/mo
Concept papers
01
X
20
=
020 hrs/mo
Utility assessments
01
X
02
=
002 hrs/mo
Justification papers
01
X
06
=
006 hrs/mo
Project summaries
05
X
05
=
025 hrs/mo
Memos for record
18
X
01
=
018 hrs/mo
Outside correspondence
15
X
02
=
=
014 hrs/mo
Investigation summary
05
X
03
=
015 hrs/mo
Cmd Operating Budget
01
X
03
=
006 hrs/mo
Impact statements
01
X
05
=
005 hrs/mo
Data manipulation*
02
X
03
=
006 hrs/mo
Misc. other
25
X
01
=
025 hrs/mo
TOTAL MANHOURS REQUIRED
=
1303 hrs/mo
-
.
..
24
-
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
X
04
020 hrs/mo
Stage essays
01
X
04
=
=
024 hrs/mo
Historical report
01
X
05
020 hrs/mo
Significant Events log
01
X
03
030 hrs/mo
Talking papers
02
X
07
=
=
=
160 hrs/mo
Proj. Ofc. rpts
05

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
. .
EXAMPLE 3: ADP MEMORY REQUIREMENTS
(Example report from 1985 System analysis - not current information)
Date of survey:
October 28, 1985
Application
# of users
# of files
# of sectors
Memory reg.
Documents
23
1,229
39,933
10,222,848 bytes
CP/M programs
6
11
10,703
2,739,968 bytes
BASIC programs
14
106
7,811
1,999,616 bytes
Indices
31
54
5,880
1,505,280 bytes
OIS data files
7
23
3,935
1,007,360 bytes
Glossaries
8
34
1,356
347,136 bytes
Message in-basket
1
1
553
141,568 bytes
GLOBAL objects
7
27
477
122,112 bytes
BASIC data files
13
16
427
109,312 bytes
BASIC indexed files
1
1
14
3,584 bytes
BASIC indexed keys
1
1
13
3,328 bytes
TOTALS
31
1,503
71,102
18,202,112 bytes
25
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1

Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
.+
EXAMPLE 4: FILING AND HISTORICAL REQUIREMENTS
(Example report from 1985 System analysis - not current information)
ITEM
File?
Length of time
Monthly Training Rpt.
Y
05 years
Daily Training Summ.
Y
05 years
File backup record
Y
.5 year
Files deleted report
Y
02 years
Operations rqmnts. pkg.
Y
10 years
Session transcripts
Y
10 years
Session reports
Y
10 years
Session results package
Y
10 years
Training database
N
N/A
Access Roster database
N
05 years
Technical summaries
Y
10 years
Inter-office corr.
Y
02 years
Proj. Ofc. rpts.
Y
10 years
Stage essays
Y
02 years
Travel reports
Y
10 years
Travel reimbursement
Y
10 years
Briefing texts
Y
02 years
Charts/diagrams
Y
as needed
Mileage report
Y
01 year
Physical training rpt
Y
01 year
Contract negotiations
Y
10 years
Contract evaluations
Y
10 years
Fact sheets
Y
02 years
Decision papers
Y
10 years
Information papers
Y
02 years
Historical report
Y
10 years
Personnel evaluations
Y
10 years
Significant Events log
Y
10 years
Staff studies
Y
10 years
Concept papers
Y
10 years
Utility assessments
Y
10 years
Justification papers
Y
10 years
Cover letters
Y
as needed
Project summaries
Y
02 years
Memos for record
Y
as needed
Outside correspondence
Y
as needed
Talking papers
Y
as needed
Investigation summary
Y
10 years
Cmd Operating Budget
Y
10 years
Proj. Ofc. historical
Y
10 years
Impact statements
Y
10 years
Misc. other
Y
as needed
26
Approved For Release 2000/08/08 : CIA-RDP96-00789R002900090001-1
....
-